前言
大家是不是已经用过一段时间的Flux.1,来出AI高画质的图片,这不SD3.5也来了吗,今天老崔就带大家来个2大头部AI绘画模型的评测,希望对你的AI绘画有帮助!
我们这次不用开源的程序(文末有惊喜!)直接在在线webUI上测试,这里我们采用的平台是https://www.liblib.art/sd
所有的AI设计工具,安装包、模型和插件,都已经整理好了,👇获取~
我们先用同一组提示词来测试,F.1和SD3.5相同参数下的出图效果。
1、A woman in a blue-green dress, with long black hair and red lips, holding an umbrella with lotus flowers around her feet. Full-body photo, front view, fantasy background, exquisite facial features, detailed depiction, fantasy art style, Chinese aesthetics, fairy tale core, light purple-pink tone, in the style of Chinese aesthetics
SD3.5出图:
F.1出图:
2、A beautiful woman wearing traditional Hanfu , with flowing hair and exquisite makeup on her face, stands in front of the Great Wall at sunset. She is dressed in gorgeous traditional costumes adorned with red silk ribbons, and she wears an orange floral headdress. Her dress features complex patterns and colors such as blue-green, dark green, and light brown. The background includes ancient buildings, towers, and a blue sky.
SD3.5出图:
F.1出图:
3、A young and beautiful Chinese woman wearing Hanfu, holding a white rabbit in her arms with the moon behind them,The photo was taken using a Canon EOS R5 camera and features volumetric lighting. It showcases high resolution, intricate details, and is presented in full body.
SD3.5出图:
F.1出图:
4、fantasy illustration of knights on horseback and in the background an enormous dragon with wings spread, flying towards camera, detailed armor, red cape flowing behind knight
SD3.5出图:
F.1出图:
5、A beautiful girl with big eyes, wearing an iridescent veil on her head and pastel colored . She has delicate features and long hair that is decorated in soft colors. The background color of the portrait should be white to highlight her face. This painting style can show a gentle atmosphere and exude feminine charm. in the style of soft colors.
SD3.5出图:
F.1出图:
通过以上F.1和SD3.5相同参数下的出图效果对比,大家可以明显看出在出真人图片的时候F.1明显要比SD3.5更真实更细腻一些,二者在生成二次元图片的时候效果差不多。
大家如果想体验一下新版SD3.5出图工作流的话,可以回复SD,取得相应的模型,大家可以在本地测试出图效果。
这里直接将该软件分享出来给大家吧~
1.stable diffusion安装包
随着技术的迭代,目前 Stable Diffusion 已经能够生成非常艺术化的图片了,完全有赶超人类的架势,已经有不少工作被这类服务替代,比如制作一个 logo 图片,画一张虚拟老婆照片,画质堪比相机。
最新 Stable Diffusion 除了有win多个版本,就算说底端的显卡也能玩了哦!此外还带来了Mac版本,仅支持macOS 12.3或更高版本。
2.stable diffusion视频合集
我们在学习的时候,往往书籍源码难以理解,阅读困难,这时候视频教程教程是就很适合了,生动形象加上案例实战,一步步带你入坑stable diffusion,科学有趣才能更方便的学习下去。
3.stable diffusion模型下载
stable diffusion往往一开始使用时图片等无法达到理想的生成效果,这时则需要通过使用大量训练数据,调整模型的超参数(如学习率、训练轮数、模型大小等),可以使得模型更好地适应数据集,并生成更加真实、准确、高质量的图像。
4.stable diffusion提示词
提示词是构建由文本到图像模型解释和理解的单词的过程。可以把它理解为你告诉 AI 模型要画什么而需要说的语言,整个SD学习过程中都离不开这本提示词手册。
5.SD从0到落地实战演练
如果你能在15天内完成所有的任务,那你堪称天才。然而,如果你能完成 60-70% 的内容,你就已经开始具备成为一名SD大神的正确特征了。
这份完整版的stable diffusion资料我已经打包好,需要的点击下方插件,即可前往免费领取!